At least if some students at the Technical University of Munich have their way of making the survival forecasts for the characters of the last season with the help of
Represent machine learning algorithms. Shortly before the start of the last season of the TV series "Game of Thrones" (GoT), students at a computer science seminar at the Technical University of Munich (TUM) embarked on an unusual scientific mission: algorithms programmed by them predict who would climb the coveted iron throne will.

Royal computer science

Students at the Technical University of Munich developed an IT application that searches the Internet for data on the "Game of Thrones" series and processes the numbers using artificial intelligence algorithms. This enables them to predict the chances of survival of the individual Game of Thrones characters. As early as 2016, shortly before the sixth season aired, students on the same course developed an algorithm that correctly predicted the resurrection of Jon Snow. The algorithm developed by the students predicts this time that Daenerys Targaryen has the highest chance (99 percent) of surviving the GoT world. The king's right hand man, Tyrion Lannister, also has a promising 97 percent survival rate.

"While the task of predicting the chances of survival for Game of Thrones characters is based on data from the world of imagination, in reality the same artificial intelligence techniques are used and have a strong influence on our everyday lives."

Lead mentor of the course, Dr. Guy Yachdav

Survival rates are predicted using longevity analysis. This technique is similar to scientific applications that study the effects of treatments and complications in cancer patients. The full list of characters and their chances of survival are online at accessible. Experienced GoT fans can familiarize themselves with the forecasting technology on the website. For example, the fact that Sansa Stark was born in Winterfell at the House of Stark and was only married once seems to increase his chances in the coming season. Your predicted probability of death is only 73 percent. Fans interested in exploring the differences between the plot of the TV series and the story from the books on which the TV series is based will find a head-to-head comparison of the details of the characters, including age, status (Dead vs. Alive) and predictions about the probability of survival.

"The combination of teaching and passion is a brilliant way to give students experience in using these new tools and prepare them to create great new applications on their own after graduation."

Professor Burkhard Rost, head of the chair for bioinformatics at TUM

The work on these algorithms is part of a JavaScript seminar that is held every semester at the Computer Science Department of the Technical University of Munich. During the course, participants will learn how to design, develop and use intelligent computer systems.

We publish the latest news from the areas of board and card games, literature and film, toys, video games and events. Send us an email with your press release.